Ad
Ad
Ad
Pages: [1] 2 3 ... 6 »   Bottom of Page
Print
Author Topic: LAB COLOR  (Read 26352 times)
tad
Newbie
*
Offline Offline

Posts: 32


« on: November 13, 2007, 11:02:19 PM »
ReplyReply

Just finished Photoshop LAB Color by Dan Margulis. I found it fascinating and was wondering if, when and why anyone employs LAB color in their workflow? I applied some of the methods to some of my landscape photos and was very impressed with the outcome.

If this belongs in the beginners forum I apologize.
« Last Edit: November 14, 2007, 09:14:18 AM by tad » Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5415


WWW
« Reply #1 on: November 13, 2007, 11:47:45 PM »
ReplyReply

It's L*, a*, b*...not LAB (it's ok to shorten to Lab but it ain't LAB).

See: Lab color space

Lab is fine for "certain" image corrections that can't be done in RGB or CMYK...but it ain't no magic bullet. Use it when and where appropriate. Also note that converting to Lab from RGB is _NOT_ lossless. There is a one time quantization error the first time you convert and Lab isn't very forgiving when working in 8 bit/channel.

And one wonders why ol' Dano loves Lab so much but calls Pro Photo RGB a "super-wide" color space and thinks it's dangerous.
Logged
pfigen
Sr. Member
****
Offline Offline

Posts: 450


WWW
« Reply #2 on: November 14, 2007, 01:48:00 AM »
ReplyReply

Well, I use Lab all the time, practically on a daily basis, particularly on provided stock images, but on my own as well. I'd say about fifty percent of what I do goes through Lab at some point. The theory put forth by so many would suggest that converting to Lab is too destructive is just that, theory. After converting hundreds if not more images, the only ones that showed any type of visible degradation were those that were screwed up to begin with. There are so many other things to worry about than a few lost llevels. As far as why Dan uses Lab and doesn't particularly care for ProPhoto RGB, you'll have to read his book. The answers are all in there, very clearly at that. Lab Color has to be the single best book in digital imaging I've ever read. Dan pushes the envelope with the way he thinks more than any other imaging author and challenges the reader to do the same for him or her self.
Logged
stewarthemley
Guest
« Reply #3 on: November 14, 2007, 04:01:51 AM »
ReplyReply

Jeez - I'm running for cover!!!
Logged
papa v2.0
Full Member
***
Offline Offline

Posts: 194


« Reply #4 on: November 14, 2007, 05:10:16 AM »
ReplyReply

or it can be written as CIELAB

it should always be   CIE 1976(L*a*b*)  or    L*,a*,b*

but it not correct to shorten it  to Lab
Logged
alanrew
Jr. Member
**
Offline Offline

Posts: 75


WWW
« Reply #5 on: November 14, 2007, 07:20:08 AM »
ReplyReply

For an alternative perspective on the 8 bit versus 16 bit debate, and Dan's position, this page on Bruce Lindbloom's site offers another view, which is worth reading even if you don't agree with it.

I think the bottom line is to use whatever workflow gives you the results that you (and your clients) want, even if the theory doesn't agree with what you're doing. OTOH, if you aren't happy with your results, be prepared to seek the knowledge of others.

FWIW, I use 16-bit ProPhoto RGB for RAW conversions, as well as subsequent editing in Photoshop. Currently, printing paths in Windows GDI are limited to 8 bits per channel, but this limit will probably be raised when the new Vista colour management gets exploited by printer driver writers. So at least I'm prepared for that.

One feature of colour management that I've discovered time & time again, over several years of learning about it, is that you can get to a point where you're happy with your workflow, and think you have a sufficient understanding of it, then along comes a new fact out of the blue that makes you realise that you don't understand it after all. Like using Adobe RGB as my PS workspace, then reading an article by Bruce Fraser a couple of years ago that reported (I'm writing from memory here as I can't find the article) that this particular space is poor at representing dark greens and browns. At this point I switched to ProPhoto RGB as well as 16 bits (because PS CS supported 16 bit layers etc.).

Regards,

Alan
Logged
wollom
Jr. Member
**
Offline Offline

Posts: 60


« Reply #6 on: November 14, 2007, 07:40:08 AM »
ReplyReply

Hey TAD,

I can't resist a response on this one.  Making a wish and getting an instant answer from God is what happened on this post. ...and SCHEWE came down and said...

I think Lab is great when you want to fiddle with with the Luminiance, the light and dark, in an image.

So I'm going to add a question, and maybe Schewe will answer me too.

When is it worth taking an image to Lab?  Seriously.

wollom  

(Tad did say he/she was a beginner, remember?)
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 8579



WWW
« Reply #7 on: November 14, 2007, 08:23:16 AM »
ReplyReply

What is useful is to see what the LAB color model is based and really used for, how large a color space it is, what happens when you convert (ugh) 8-bit RGB to Lab and back, plus the time it takes and when you can avoid all these issues by using the lumonisty blending modes in Photoshop.

Lab certainly has a place in image processing. But its been way over blown. Its primarily become the poster child for one author who, as Jeff said, thinks its super cool but has no issues dismissing ProPhoto RGB as 'dangerous'. So what he says should often be taken with a grain of salt.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
digitaldog
Sr. Member
****
Offline Offline

Posts: 8579



WWW
« Reply #8 on: November 14, 2007, 08:25:20 AM »
ReplyReply

Quote
FWIW, I use 16-bit ProPhoto RGB for RAW conversions, as well as subsequent editing in Photoshop. Currently, printing paths in Windows GDI are limited to 8 bits per channel, but this limit will probably be raised when the new Vista colour management gets exploited by printer driver writers. So at least I'm prepared for that.

A little OT but OS X 10.5 apparently has this functionality. Jeff, can we talk about what's coming from you know who, the printer company (and the need for application support for 16-bit printing)?
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
tad
Newbie
*
Offline Offline

Posts: 32


« Reply #9 on: November 14, 2007, 08:30:56 AM »
ReplyReply

I apologize profusely for my obvious ignorance and the incorrect use of "LAB" in my request for further knowledge in image processing. I can only hope for forgiveness and that the moderators allow me to remain an active participant on this forum.

In the future, if I am still allowed here, I will remain in the beginners section until my vocabulary and word usage is up to par with that of the more senior members.

Always grateful,

Tad
« Last Edit: November 14, 2007, 08:35:57 AM by tad » Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 8579



WWW
« Reply #10 on: November 14, 2007, 08:40:15 AM »
ReplyReply

Quote
The theory put forth by so many would suggest that converting to Lab is too destructive is just that, theory. After converting hundreds if not more images, the only ones that showed any type of visible degradation were those that were screwed up to begin with.

I would agree that all image correction involves data loss. One issue I have with Dan is that:

A. He dismisses this, especially high bit workflows when he (or we) never know when an image will break.

B. The output is rarely if ever defined. You can wonk on an 8-bit image going to a halftone dot till the cows come home and never see the effect (everything to Dan is this output) and really hose output to a fine ink jet or Contone printer.

C. It takes time to do the conversions on big files and often, you can produce the same effects visually in less time, with more post processing control using blending modes.

As for the data loss, its a fact, not a theory. Here's a test Richard Wagner did that I believe he tried or maybe actually got through the heavily moderated ACTL of Dan's:

Quote
OK, some preliminary results. All image conversions were done in PS
CS2. Where appropriate, conversions were made using RelCol and BPC,
with no dither. Image level counting was done using Levels 1.2,
©2001 Bruce J. Lindbloom, running under Classic 9.2 on Mac OS
10.4.7. The test file was the 8-bit 4096 x 4096 pixel image
containing all unique colors provided by Bruce Lindbloom via his web
site. http://www.brucelindbloom.com/index.html?RGB16Million.html
This is a good file to start with, as it does not contain JPEG
artifacts and it has not been previously subjected to profile
conversions that might bias the results.

Original Image:
File Name: RGB16Million.tif
Image Dimensions: 4096 x 4096 = 16,777,216 pixels
Color Type: RGB
File Format: TIFF
Bits/Channel: 8
Color Space: sRGB (assigned in PS CS2)
There are 16,777,216 unique colors in this image.
==========

Convert the above image to Lab:
There are 2,186,765 unique colors in this image.
==========

Convert the Lab image back to sRGB:
There are 2,186,295 unique colors in this image.

Note that this data matches very closely what is presented on Bruce
Lindbloom's site. I'm not sure why the slight difference - changes
in the CMM? Also, as he stated, "back and forth" trips to Lab do not
result in increased image degradation. It is primarily the first
trip to Lab that results in quantization loss. I have confirmed
this. (additional data not shown.)
==========

OK, how about Adobe to Lab? Better than sRGB, as I predicted in a
previous post.
There are 3,135,822 unique colors in this image.
==========

And ProPhoto to Lab - should be better yet - and it is. As expected,
moving from a wide-gamut color space to Lab results in less
quantization loss than moving from sRGB to Lab.
There are 6,236,954 unique colors in this image.
==========

In testing some of my own scanned wildlife images, the quantization
loss was frequently around 50% - 70% when converting from AdobeRGB to
Lab. For example, a 14-bit scan of a Kirtland's Warbler began with
18,617,959 unique colors; conversion to 8-bit resulted in 513,680
unique colors, and conversion of that image to Lab resulted in
279,760 unique colors.

Is any of this significant? In looking at the images, it doesn't
appear to be, although the same can be said for many other operations
that result in data loss. It is not clear what effect this may have
on an image if many operations follow. Avoiding the transition to
Lab avoids the data loss, although this obviously gives up the
advantages tof techniques unique to Lab.

--Rich Wagner
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
tad
Newbie
*
Offline Offline

Posts: 32


« Reply #11 on: November 14, 2007, 09:12:38 AM »
ReplyReply

I understand and am quite used to people wanting to pontificate on a web forum, but what would be really nice is if someone would just actually answer my original question.
Logged
DarkPenguin
Guest
« Reply #12 on: November 14, 2007, 09:29:55 AM »
ReplyReply

Quote
Just finished Photoshop LAB Color by Dan Margulis. I found it fascinating and was wondering if, when and why anyone employs LAB color in their workflow?

I dont.  Never.  Because I get the results I want without it.
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 8579



WWW
« Reply #13 on: November 14, 2007, 09:41:40 AM »
ReplyReply

Quote
I dont.  Never.  Because I get the results I want without it.
[a href=\"index.php?act=findpost&pid=152721\"][{POST_SNAPBACK}][/a]

Ditto. I do all the heavy lifting of tone and color rendering in a Raw converter. As such, I'd say 98% of all global work is done, I use Photoshop for local corrections and as yet, haven't seen any reason to convert to Lab for that. Hope that answer the OP question.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Schewe
Sr. Member
****
Offline Offline

Posts: 5415


WWW
« Reply #14 on: November 14, 2007, 10:21:14 AM »
ReplyReply

Quote
When is it worth taking an image to Lab?  Seriously.
[a href=\"index.php?act=findpost&pid=152688\"][{POST_SNAPBACK}][/a]

When you need to work on the color separately from the luminance or you need to adjust the contrast of the color separately from the luminosity. The ab channels (not unlike HSL or HSB) give a totally different impact when using Photoshop's tool set...

On the other hand, I would never really go into Lab for the purpose of working only on the luminosity as I've found the L channel so close to the luminance blend mode that it ain't worth the conversion to Lab.
« Last Edit: November 14, 2007, 10:24:15 AM by Schewe » Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 8579



WWW
« Reply #15 on: November 14, 2007, 10:33:37 AM »
ReplyReply

Quote
I'd like to hear some focused discussion on this as well. 

Take the example of steepening the a/b curves (or is that a*/b*, or A, B or ...) to "enhance color" (sorry, that's probably not the "correct" term either).  Say you even do this on a copy image that you selectively reveal, layered in your "original" 16bit ProPhoto file.   What is a better way to accomplish this manipulation without going to... let me see... L*, a*, b*?
[a href=\"index.php?act=findpost&pid=152734\"][{POST_SNAPBACK}][/a]

I'd use Vibrance in Lightroom. This would be totally non destructive, instruction based rendering. I could make alterations to this (and iterations) without taking up more than a few k's of disk space.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #16 on: November 14, 2007, 01:34:28 PM »
ReplyReply

Quote
I'd use Vibrance in Lightroom. This would be totally non destructive, instruction based rendering. I could make alterations to this (and iterations) without taking up more than a few k's of disk space.
[a href=\"index.php?act=findpost&pid=152738\"][{POST_SNAPBACK}][/a]
Vibrance is not nearly as flexible. Working with the a/b curves you can steepen one more than the other, you can also use blend-if to limit the effect from colors that are already saturated. I do this to some extent for almost every landscape/nature image I postprocess, it's an invaluable tool in my workflow. I also sharpen the L channel while I'm there, although I wouldn't make the trip to Lab just for that.
Logged

digitaldog
Sr. Member
****
Offline Offline

Posts: 8579



WWW
« Reply #17 on: November 14, 2007, 01:42:43 PM »
ReplyReply

Quote
Vibrance is not nearly as flexible. Working with the a/b curves you can steepen one more than the other, you can also use blend-if to limit the effect from colors that are already saturated. I do this to some extent for almost every landscape/nature image I postprocess, it's an invaluable tool in my workflow. I also sharpen the L channel while I'm there, although I wouldn't make the trip to Lab just for that.
[a href=\"index.php?act=findpost&pid=152785\"][{POST_SNAPBACK}][/a]

Look, I have no issues rendering global tone, color and saturation (among other things) from Raw to produce the color appearance I want to represent. I don't need to use a blend if and jump through all kinds of hoops. Just because you have layers, blend modes and so forth, doesn't mean you have to go through a 38 step, convoluted process. You might, it may be 3 steps but as yet, I haven't found the need. I like to practice KISS.

There's no reason to sharpen the L channel either, another exercise in time loss and data loss. Just run the sharpening and fade luminosity. Is it 100% identical? No but it addresses the problem you're going after in Lab, color fringing. Its faster, it causes a lot less data loss AND you have the opacity slider to boot.

Lab has a role in image processing. But as I said, its recently become the big macho color space for doing anything and everything according to Dan. If he would spend just a little time looking at Raw rendering (instead of slamming it or trying to teach how to polish turds using Photoshop), his ideas might be easier to swallow. He sees everything from the perspective of a hammer in which every image correction (NOT rendering) is a nail.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6818


WWW
« Reply #18 on: November 14, 2007, 02:02:37 PM »
ReplyReply

Before LR/CR4.1, I found it sometimes worthwhile converting an image to L*a*b for enhancing colour separation and saturation. It is also correct that the Blend IF sliders worked on the *a* and *b* channels in Layer Styles help very much to target and control the impacts. (But for luminance work, as observed above, one is just as well off using the RGB composite curve in Luminosity Blend Mode). At the same time, it must be said, this L*a*b is not an easy space to work with - takes alot of experience learning to control it and to predict what various curve moves will do, because it is anything but intuitive. It is also inconvenient making trips back and forth between colour spaces because Adjustment Layers cannot be preserved, so the workflow needs to be organized around trips to Lab in a way that avoids that inconvenience. (Here I'm ignoring the loss of information associated with colour space exchanges because that has already been well demonstrated and thoroughly discussed.) Hence all these very obvious and practical factors taken into account, if there are easier, non-destructive ways of achieving the same results without exchanging colour spaces, so much the better. LR and CR4.x have vastly increased this capability, not only because of the Vibrance and Clarity controls in the Basic Tab, but also because of the 24 independent targeted controls one has in the HSL Tab for eight colour groups of significant interest to photographers. To sum-up a short answer to the OP's question: do as much as the new (and old) tools allow in LR/CR4.x and CS3 before moving to L*a*b, and if you still find you need a particular effect just not otherwise achievable, go ahead and try L*a*b, but be mindful of the workflow implications.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
David WM
Full Member
***
Offline Offline

Posts: 241


WWW
« Reply #19 on: November 14, 2007, 10:20:32 PM »
ReplyReply

After reading the book I found Lab to be a very useful tool, I can do things that I cant seem to achieve in other colour spaces.  These mainly involve modifying specific colours or help in making selections in an image. I also use the L chanel for contrast adjustments when I don't want to alter colours (could use luminance layer, but isn't adding a layer about the same amount of work as switching spaces?) I understand it is just a part of the whole range of tools available in photoshop and as always there are a number of ways to do a lot of things, and which method is "preferred" is subjective or at least the subject of sometimes rigorous debate .  I found some of the workflows in Dan's books very practical, its a pity he has to waste so much text space  defending his methods, after all diversity is good!  I am not a lightroom user so I am not sure how that would affect my opinions. I am looking forward to receiving RWCR CS3 so I can get up to speed on what Camera Raw has to offer. I think Dan's book is about utilising photoshop for image adjustments, not about what could be achieved in other programs like Lightroom. Its a very detailed book so I can't imagine coming to grief applying the methods if you understand the text, but it does take a while to get used to it.

David

Quote
I found it fascinating and was wondering if, when and why anyone employs LAB color in their
[a href=\"index.php?act=findpost&pid=152622\"][{POST_SNAPBACK}][/a]
Logged
Pages: [1] 2 3 ... 6 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad